A 200GHz 200-Pixel 2D Near-Field Imager for Biomedical Applications

Abstract: 

Near fields, unencumbered by the restrictions imposed by diffraction limits, can be leveraged in high-resolution sub-wavelength imaging systems [1]. Utilizing the near-fields generated by arrays of mmWave and THz resonators can be a compelling candidate for realizing such systems. On the one hand, by increasing the frequency of operation, the resonators' physical footprint shrinks, enhancing the imaging resolution. On the other hand, their highly localized and easily perturbable EM-fields and narrow-band nature can potentially bring about high SNR detection and characterization of samples as small as single cells [2]. These potentials, however, have never been simultaneously realized due to reasons such as: 1) The challenges in the generation and the detection of mmWave and THz signals have limited previous efforts to a single-source and/or single-detector per pixel architectures [1]–[4], wherein the imaging resolution does not scale consistently with the operation frequency. 2) The absence of a proper method for the selection and the readout of a single resonator in isolation from the rest of the array further exacerbates the resolution problem. 3) The resonator Q degrades at high frequencies and in the presence of lossy biological samples, ultimately reducing the sensitivity limits. This work advances the state of the art by overcoming the aforementioned challenges through several architectural innovations: 1) Q-boosted Active Split-Ring Resonators (ASRR) are introduced, which allow dense integration and individual pixel selection with minimum capacitive loading to preserve the sensitivity. 2) ASRRs can be selectively activated and deactivated, which enables using a single source and a detector for multiple pixels. 3) Phase detection replaces the standard power or frequency detection methods, which are prone to the remarkably high phase noise of free-running high-frequency VCOs. 4) By periodically switching the selected pixel between reference and sensing states, correlated-doubling sampling (CDS) can be used to suppress the detection chain's low-frequency noise. Combining these techniques results in a true 2-D scalable imager capable of delineating tissue cells at unprecedented sensitivity and contrast. The versatility of this platform allows it to be deployed in a plethora of applications, such as intraoperative imaging [5] and discovering novel therapeutics via autonomous real-time monitoring of cell proliferation under chemical or electrical stimuli [6].

Publication date: 
January 1, 2025
Publication type: 
Conference Paper